A recent report from the OECD, Developing Vocational Education and Training with Artificial Intelligence, examines a different aspect of AI in VET from the increasingly common topics of student use, assessment integrity and/or AI-supported teaching. Its focus instead is how AI can help governments, industry bodies and VET providers develop and update occupational standards, qualifications and curricula.
The report draws on a survey of 25 countries and detailed case studies from ten countries (Croatia, England, Estonia, Germany, Ireland, Korea, Lithuania, Mexico, the Netherlands and Switzerland). Its starting point is that conventional VET-development processes are generally collaborative and robust, but can be slow and resource-intensive at a time when occupational requirements are changing rapidly.
AI can potentially assist at several stages of the development process. It can analyse large volumes of job advertisements and labour-market information, identify emerging skills, compare occupational standards and curricula, map competencies across different qualifications, synthesise stakeholder feedback, and assist with drafting or checking learning outcomes. These uses could help VET systems identify gaps earlier and reduce some of the administrative work involved in reviewing training products. Importantly, the report presents AI as a way of supporting expert work rather than replacing the industry consultation and professional judgement on which credible vocational qualifications depend.
Current use remains uneven and mostly experimental. Public VET agencies and industry bodies are generally further advanced than individual providers or teachers. Examples include England’s SkillsCompass, which combines occupational standards, labour-market data and foresight information; a qualification-development platform in the Netherlands; and occupation-specific models in Switzerland’s ICT sector. Australia is identified among the countries using AI to help organise consultation and synthesise stakeholder contributions, while Korea, New Zealand and the Netherlands are using it in areas such as drafting learning outcomes and checking alignment between curricula and standards. The report also notes that Future Skills Organisation “is undertaking exploratory research to create potential efficiencies in using AI in training package development processes enabling faster time to market and stronger industry alignment.”
The report is careful not to present AI as a solution to every problem. Barriers include poor or fragmented data, limited AI and data literacy, a lack of secure and fit-for-purpose tools, and insufficient institutional guidance. Smaller providers and industries may be particularly disadvantaged. There are also risks that AI-generated outputs could contain errors or bias, that responsibilities for decisions could become unclear, or that consultation could be weakened if automated summaries substitute for genuine engagement. Unpublished training-product information, employer data and other commercially sensitive material also create privacy and security risks.
Against the backdrop of these developments across the OECD, it should also be noted that in Australia, ASQA has also just released its 5 Principles for the Responsible Use of AI in vocational education and training (VET).
They are “designed to support the safe, ethical and compliant use of AI across all aspects of a provider’s operations. Each principle is supported by a description and self-assurance questions which, in conjunction with the case studies, are designed to help providers navigate their adoption of AI tools and systems.”
It is also important to note that the principles do not introduce new regulatory requirements but are designed to “provide a structured way to interpret, implement and oversee AI use within providers’ existing obligations.”
